The growing sophistication and rapid proliferation of malicious software, also known as malware, presents an ever-increasing security threat to personal and enterprise computer systems worldwide. New types of malware emerge daily and spread rapidly through the Internet and local area networks, e-mail, Instant Messaging and file sharing services and other data communication technologies. Known malicious software can be automatically detected by anti-malware programs and classified into one of several categories, such as viruses, worms, Trojan horses and spyware, based on the software code or behavior pattern. However, there are many other types of malware with hidden code and constantly changing behavior, such as polymorphic viruses and obfuscated malware, which make automatic detection and classification difficult. As a result, the anti-malware programs may fail to automatically detect and classify these types of malware or spend such a significant amount of time and system resources on the analysis of these programs that the detection process becomes inefficient. Accordingly, there is need for new and more effective methods for automatic detection and classification of malicious software.